Volcanic Event Recurrence Rate Model (VERRM): Incorporating Radiometric Ages, Volcanic Stratigraphy and Paleomagnetic Data into a Monte Carlo Simulation to Estimate Uncertainty in Recurrence Rate through Time

Abstract:

Traditional methods used to calculate recurrence rate of volcanism, such as linear regression, maximum likelihood and Weibull-Poisson distributions, are effective at estimating recurrence rate and confidence level, but these methods are unable to estimate uncertainty in recurrence rate through time. We propose a new model for estimating recurrence rate and uncertainty, Volcanic Event Recurrence Rate Model. VERRM is an algorithm that incorporates radiometric ages, volcanic stratigraphy and paleomagnetic data into a Monte Carlo simulation, generating acceptable ages for each event. Each model run is used to calculate recurrence rate using a moving average window. These rates are binned into discrete time intervals and plotted using the 5^th, 50^th and 95^th percentiles. We present recurrence rates from Cima Volcanic Field (CA), Yucca Mountain (NV) and Arsia Mons (Mars). Results from Cima Volcanic Field illustrate how several K-Ar ages with large uncertainties obscure three well documented volcanic episodes. Yucca Mountain results are similar to published rates and illustrate the use of using the same radiometric age for multiple events in a spatially defined cluster. Arsia Mons results show a clear waxing/waning of volcanism through time. VERRM output may be used for a spatio-temporal model or to plot uncertainty in quantifiable parameters such as eruption volume or geochemistry. Alternatively, the algorithm may be reworked to constrain geomagnetic chrons. VERRM is implemented in Python 2.7 and takes advantage of NumPy, SciPy and matplotlib libraries for optimization and quality plotting presentation. A typical Monte Carlo simulation of 40 volcanic events takes a few minutes to couple hours to complete, depending on the bin size used to assign ages.